HIV-1 Wild Type
Protease (PBD ID: 5UPZ) from Human Immunodeficiency Virus-1
Created by:
Rachel Mellon
Infection of
Human Immunodeficiency Virus-1 (HIV-1) from the Lentivirus genus leads to
chronic weakening of the immune system, leading to Acquired Immune Deficiency
Syndrome (AIDS). Infection is usually due to transfer of bodily fluids between
individuals. AIDS makes a patient more susceptible to opportunistic infections
that do not usually cause illness or death in a non-immunocompromised
individual (1). In an infected individual, HIV-1 enters memory T-cells by
binding the external glycoprotein gp120 to the CD4+ receptor found on the
membrane of memory T-cells. Once inside the host cell, HIV-1 undergoes
replication by using its own reverse transcriptase to insert itself into the
host cells genome. HIV-1 takes over the function of the T-cell, using both
HIV-1 and host cell transcriptional machinery to create new viral particles.
The viral particles leave the T-cell using the vesicular sorting pathway via
binding TSG101 of Gag. The poly protein Gag-pol is cleaved by HIV-1 protease (PBD ID: 5UPZ) (2). This cleavage allows
HIV-1 particles to assemble and maturate the protein coat. It is only when the
HIV-1 particles are maturated they are pathogenic and can infect other T-cells
(3).
HIV-1 exists as quasispecies, and it is one of the most
polymorphic viruses known. In recent years HIV-1 is showing an increased
resistance to drugs. Therefore, it is critical to work towards understanding
the mechanism of HIV-1 pathogenesis to prevent infection (1). One way to
develop a wider range and better targeting drugs is to specifically focus on
the HIV-1 protease and study the crystal structure with novel HIV-1 protease
inhibitors such as GRL-0518A. Inhibiting HIV-1 protease prevents the HIV-1
particle from protein coat maturation, resulting in the production of
non-infectious viral particles, hence HIV-1 protease is an ideal target for
antiviral therapy (3).
In physiological conditions, HIV-1 protease exists as a 22480.68 Da dimer of two identical subunits with two fold (C2) symmetry (4, 5). The molecular weight of the protein was calculated using ExPASy, a bioinformatics server which provides a range of software tools and scientific databases for proteins or genes of interest. Containing 99-amino acid residues in each monomer, each subunit contains one α-helix, approximately eight random coils and nine β-sheets (5). The dimeric structure confers stability by the NH2 terminal residues 1-4 and COOH terminal residues 96-99 forming a four stranded β-sheet. The four stranded β-sheet of the dimeric interface results in polar residues Gln-2, Thr-4, Thr-96 and Asn-98 facing towards the solvent, and packing of the hydrophobic residues Pro-1, Ile-3, Leu-97 and Phe-99 to the interior of the dimer interface, promoting inter-chain interactions (6). Zhang reports this stable dimeric structure contains 4 salt bridges and 34 hydrogen bonds. As a typical hydrogen bond is approximately -30kJ/mol, the total strength of hydrogen bonds in this small region of the enzyme would be ~1020 kJ/mol (7). Todd has suggested through thermodynamic measurements that the most stable part of the mature HIV-1 protease is the terminal β-sheet (8). The isoelectric point of the HIV-1 protease dimer is 9.66, indicating an abundance of basic residues on the protein surface which will be positively charged at physiological pH (4).
Upon dimerization of the HIV-1 protease monomers, the active site is
formed with each monomer contributing one of the two aspartic acid residues.
Therefore, it is only when HIV-1 protease is in this dimeric form that the
protease activity is functional. HIV-1 protease is also categorized as a
catalytic aspartate protease as HIV-1 protease due to these aspartic acid
residues in the active site. This dimer functionality is specific to retroviral
proteases, as cellular aspartic proteases are active as monomers (3). The dimeric active site is composed of a highly conserved triad of
Asp-(25, 25’), Thr-(26, 26’) and Gly-(27, 27’) which forms a hydrogen bond
network often referred to as the “fireman’s grip” (6). Asp-25 and Asp-25’ are
planar in the HIV-1 protease dimer and directly interact with substrates and
inhibitors. Researchers been hypothesised that the function of Thr-26 and
Thr-26’ is to stabilize the active sit by forming hydrogen bonds across the
dimeric interface, and the function of Gly-27 and Gly-27’ is to bind to the
substrate or inhibitor in an orientation that favours the carboxylate groups
from Asp-25/Asp-25’ to attack the peptide bond of a substrate or inhibitor (9).
HIV-1 protease also contains“flaps” from residues 47-56. These flaps are flexible and are presumed to
play a significant role in substrate and inhibitor binding, as they adopt a
closed conformation when a substrate or inhibitor is bound to HIV-1 protease
(6).
HIV-1 protease (PDB ID: 5UPZ) was crystallized using hanging drop vapour
diffusion and structural data was obtain by X-ray diffraction using a single
wavelength of light. HIV-1 protease contains four ligands. Glycerol (PDB: GOL),
Na+ and Cl- are ligands presumed to be used to
crystallize the enzyme. The fourth ligand, N~3~-{(2S,3R)-3-hydroxy-4-[{[4-(hydroxymethyl)phenyl]sulfonyl}(2-
methylpropyl)amino]-1-phenylbutan-2-yl}-N~1~-
methyl-N~1~-[(4-methyl-1,3-oxazol-2-yl)methyl]benzene- 1,3-dicarboxamide,
otherwise known as GRL-0528A (PDB ID: 8HD) is an inhibitor of HIV-1 protease
activity (5). GRL-0528A has a structure similar to darunavir, a potent drug
used to treat HIV-1 quasispecies. Similar to darunavir, GRL-0528A functions by forming a hydrogen
bond from the urethane NH with the carbonyl oxygen found on Gly-27. The
carbonyl oxygen and sulphonamide oxygen of GRL-0528A interact with the amide
moieties of Ile-50 and Ile-50' located in the flaps of HIV-1 protease. In
addition, the P2-isophthalamide carbonyl group of GRL-0528A forms a hydrogen
bond with the side chain and backbone amide of Asp-29, found just outside of
the catalytic center. The inhibitor is also stabilized in the HIV-1 protease by hydrophobic interactions of the P2-isophthalamide group wedging into a space
surrounded by Ile-50’, Ala-28, Val-32, Ile-47 and Ile-84 as well as the
P3-4-methyl-2-oxazole group interacting with the side chain of Val-82’.
However, a disadvantage of the GRL-0528A inhibitor is that it cannot form a
hydrogen bond to Gly-27 in the active site if the oxazole ring is flipped
towards the S2-extended site (10).
The amino acid sequence of HIV-1 protease has been
examined using PSI-BLAST to search protein databases to find sequence
similarities to HIV-1 protease. PSI-BLAST assigns an E value based on the
homology of both sequences and by amino acids which are present in the subject
sequence, but not the query, known as gaps. Therefore, sequence homology
decreases the E value, where-as gaps increase the E value. An E value less than
0.05 is considered significant (11). The Dali Server has also been used to
examine HIV-1 protease to find proteins with similar tertiary structures. The
Dali Server assigns a Z-score using a sum-of-pairs method to compare the
intramolecular distances between two proteins. A Z-score above 2 indicates that
a protein has similar folds to HIV-1 protease (12). Using both of these
software, it was found that Xenotropic Murine Leukaemia Virus-related Retrovirus Protease (PDB ID: 3NR6) from Xenotropic Murine
Leukaemia-related Retrovirus (XMRV) has an E value of 2e-31 and a
Z-score of 9.6 (11, 12). XMRV protease contains a highly similar protein sequence and tertiary structure to HIV-1 protease. XMRV has recently moved from a mouse host to a
human host as XMRV have lost their envelope proteins no longer have a receptor
on murine cells. XMRV has been shown to be present in chronic fatigue syndrome
and prostate cancer patients, yet it does not appear to contribute to the
development of AIDS (13). XMRV protease functions similarly to HIV-1 as it
cleaves Gag-pol to maturate the protein coat. Therefore, XMRV protease is also
an important protein to study as it leads to disease and it functions in a
similar way to HIV-1 protease. XMRV protease also contains two identical
subunits which form a homodimer with twofold symmetry and ordered flaps which
resemble the open conformation of HIV-1 protease. However, the chain length of
each monomer is 132-amino acid residues opposed to 99-amino acid residues in
HIV-1 protease. The NH2 terminal, a partially helical secondary
structure, is present before the β1-strand,
and therefore there is no linkage of the NH2 terminus and COOH terminus by four β-strands as found in HIV-1 protease. The dimer interface of
XMRV protease is composed of hairpin loops formed by β10 and β11 strands near
the COOH terminus, where-as HIV-1 protease creates a dimer through Asp-25 and
Asp-25’ residues. The NH2 and COOH termini of XMRV protease are
longer than is found in other retroviruses, such as HIV-1 protease (14).
Furthermore, the β-strands in XMRV protease consist of alternating hydrophobic
and hydrophilic residues, in HIV-1 protease they are predominantly hydrophobic
residues (5). It has not been determined how the differences between secondary
structure found between HIV-1 protease and XMRV protease directly effect the
mechanism of function (14).
HIV-1 protease is biologically significant as it functions to cleave the poly protein Gag-pol which allows for HIV-1 particles to maturate and therefore become infectious (3). HIV-1 is showing an increased resistance to drugs, so it is especially important to study the crystal structure HIV-1 protease to enable the development of better targeting drugs which that inhibit the activity of HIV-1 protease thus producing non-infectious particles (1, 3).